It is known that 9α-hydroxy-steroids are widely used in therapy, as for example the 9α-hydroxy derivatives of pregnane steroids have glucocorticoid activity as well as the 9α-hydroxy derivatives of androstane derivatives are used as active ingredients of anti-androgen and anti-estrogen drugs.
Those 9α-hydroxy-steroids which do not have substituent in position 11 can easily be dehydrated by known chemical methods and the so obtained 9(11)-dehydro-steroids are important intermediates in the synthesis of compounds possessing high biological activity. Such compounds are for example hydrocortisone (chemical name: 11β,17,21-trihydroxy-pregn-4-ene-3,20-dione) and prednisolone (chemical name: 11β,17,21-trihydroxy-pregn-1,4-diene-3,20-dione) having anti-inflammatory activity or eplerenone (chemical name: 9α,11α-epoxy-17β-hydroxy-3-oxo-pregna-4-ene-7,21-dicarboxylic acid gamma lactone) the latter having abroad indication profile; for example it decreases the risk of mortality caused by heart and blood-vessel problems similarly to beta-blockers and it is also used for the treatment of high blood pressure and as a diuretic.
The members of the Δ4-3-keto-pregnane family were first hydroxylated in the 9α-position by Hanze and coworkers in 1958 using Cunninghamella and Helicostylum thread fungus strains (see: U.S. Pat. No. 3,038,913). In 1960 Sih and coworkers described the 9α-hydroxylation of steroids using microorganisms having Δ1-dehydrogenase enzyme activity in the presence of Δ1-dehydrogenase inhibitor (see: U.S. Pat. No. 3,065,146). Two years later Sebek carried out the 9α-hydroxylation of steroids by using Ascochyta linecola strain (see: U.S. Pat. No. 3,116,220).
In the previously mentioned patent Sih and coworkers also listed mycobacteria strains as microorganisms having steroid 9α-hydroxylase enzymes (see: U.S. Pat. No. 3,065,146).
It is known, that in 1977 Frederick and coworkers produced a new mycobacterium strain by mutagenic treatment, which only partially degraded the examined sterol substrates and therefore the 9α-hydroxy derivatives were accumulated (see: U.S. Pat. No. 4,029,549). Wovcha used the same strain—Mycobacterium fortuitum NRRL B-8119 strain—for the synthesis of new 9α-hydroxy derivatives (see: U.S. Pat. No. 4,035,236).
Wovcha and coworkers studied the role of Mycobacterium fortuitum ATCC 6842 strain in the degradation of the steroid backbone. They found that the key step in the degradation into carbon dioxide and water is the subsequent functioning of Δ1-dehydrogenase and 9α-hydroxylase enzymes. The two conversion steps can be interchanged that is both reaction steps can be carried out by two-two enzymes (group of enzymes); for example one of them, the Δ1-dehydrogenase enzyme converts the starting material and the other Δ1-dehydrogenate the 9α-hydroxy derivative. In their experiments they supposed, that the above enzymes can be induced; and by using different inducers the amount and the composition of the formed products varied significantly [Biochimica et Biophysica Acta 574, 471-479 (1979)].
In 1981 Marsheck and coworkers carried out the 9α-hydroxylation of steroid compounds by using a new mutant Nocardia canicruria strain in such a way that it was not necessary to use Δ1-dehydrogenase enzyme inhibitor. In the above mentioned examples the synthesis of 9α-hydroxy-ketolactone (chemical name: 9α,17-dihydroxy-3-oxo-17α-pregna-4-ene-21-carboxylic acid gamma lactone) is described starting from ketolactone (chemical name: 17-hydroxy-3-oxo-17a-pregna-4-ene-21-carboxylic acid gamma lactone); using 0.5 g/dm3 concentration of the ketolactone substrate the desired hydroxylated product was formed in 30% conversion (see: U.S. Pat. No. 4,397,947).
Among others Mutafov and coworkers studied the inducibility of the steroid 9α-hydroxylase enzyme using Rhodococcus sp. strain and found that 9α-hydroxy-4-androstene-3,17-dione formed as a product was a very poor inducer, since using it as an inducer slowed the reaction rate in half and the amount of the formed 9α-hydroxy product was a quarter of that when 4-androstene-3,17-dione was used as inducer [Process Biochemistry 32 (7), 585-589 (1997)].
Brzostek and coworkers studied the degradation of the steroid backbone on gene level and found that blocking the Δ1-dehydrogenase enzyme activity, which is needed for the synthesis 9α-hydroxy steroid derivatives, is difficult, because there are not only different types of Δ1-dehydrogenase enzymes but the genome contains in some cases five Δ1-dehydrogenase genes [Microbiology 151, 2393-2402 (2005)].
It is known that a microbiological step is carried out besides the chemical reaction steps in the synthesis of eplerenone, among others the hydroxylation of a valuable intermediate, the canrenone (chemical name: 17-hydroxy-3-oxo-17α-pregna-4,6-diene-21-carboxylic acid gamma lactone), was carried out with microorganisms (Diplodia, Aspergillus, Absidia sp.) (see: US Patent Applications No. 2004/087562 and 2004/097475 and further PCT International Patent Application No 2005/000865).
Another synthesis of eplerenone can be carried out via 9α-hydroxylation of canrenone. The 9α-hydroxy-canrenone (chemical name: 9α,17-dihydroxy-3-oxo-17α-pregna-4,6-diene-21-carboxylic acid gamma lactone) was first synthesized by Ng and coworkers by microbiological hydroxylation (see: PCT International Patent Appl. No 97/21720 and Hungarian patent No 222,453) and described in Example 17 in the above mentioned patents.
The 9α-hydroxy-canrenone as product is first described in patent claims in 1998 in the patent of Ng and coworkers (see: PCT International Patent Appl. No 98/25948; or later U.S. Pat. No. 7,129,345).
The international search examining authority found 18 independent inventions in the PCT-patent applications No 97/21720 and 98/25948, therefore they suggested to the inventor to file selection patent applications. As a result of this more than 100 patent applications were filed, from which several contains Example 17 of the above mentioned patent No. WO-97/21720 (see PCT No 2005/239761).
The above mentioned Example 17 describes the screening data of 83 microorganisms, which potentially have steroid 9α-hydroxylating enzyme activity and gives the TLC, HPLC/UV and LC/MS data of the products formed during the bioconversion of canrenone. From the table given there it can only be seen that the 9α-hydroxy-canrenone can be detected by the above mentioned analytical methods in the possible products or not. There is a mycobacterium in this table, Mycobacterium fortuitum ATCC 6842 strain, but there are no analytical data given in the appropriate columns. The bioconversion ability of this strain is known from the literature [publications starting from 1936, Acta Med. Rio de Janeiro 1,1], therefore it can be supposed that decomposition of canrenone took place (see U.S. Pat. Nos. 4,029,549 and 4,035,236).
This presumption is supported by a publication, which was written in 2003 by the microbiologist inventors of the above mentioned patent family. This publication contains the same table, but the Mycobacterium fortuitum strain in this table a variant of the above, developed for 9α-hydroxylation of steroids: registry number NRRL B-8119 [J. Nat. Prod. 66, 350-356 (2003)]. In this case according to the authors the Mycobacterium fortuitum NRRL B-8119 did not produce hydroxy or dehydrogenated product and there was no metabolism.
In the above mentioned Example 17 there are 3 types of Nocardia strains, namely: Nocardia aurentis, Nocardia cancicruria and Nocardia coralline strains. According to TLC and HPLC measurements the conversion products of two strains are similar to 9α-hydroxy-canrenone, but the formation of 9α-hydroxy-canrenone was disclosed by LC/MS analysis.
The only microbiological synthesis of 9α-hydroxy-canrenone in which numerical data are given is described in the above mentioned publication: Corynespora cassiicola ATCC 16718 strain was used in aerobic fermentation carried out in a flask, using 0.1 g/dm3 concentration of canrenone substrate the desired hydroxylated product was formed in 30% conversion [J. Nat. Prod. 66, 350-356 (2003)].
As it can be seen from above mentioned publications there is no such microbiological synthesis of 9α-hydroxy derivatives of canrenone or ketolactone, which is industrially applicable.
The aim of our invention is therefore to elaborate an industrially applicable microbiological process for the hydroxylation of steroid of the formula (II), wherein the meaning of -A-A′- is —CH2—CH2— or —CH═CH— group, as substrate in position 9 without considerable degradation and by-product formation.
In our initial experiments mycobacterium strains proved to be the most suitable for the microbiological synthesis of 9α-hydroxy-canrenone. The conversion ability of 38 mycobacterium and Nocardia strains were screened using ketolactone and canrenone as substrate. Among these strains there were wild type sterol degrading ones, for example Mycobacterium fortuitum ATCC 6842, or partially backbone degrading Mycobacterium fortuitum NRRL B-8129; as well as several strains, definitely developed for 9α-hydroxylation: Mycobacterium fortuitum NRRL B-8119, Mycobacterium sp. NCAIM 1072, Mycobacterium sp. NCAIM 324.
During the screening we found 3 strains, which—according to TLC analysis—produced detectable amount of 9α-hydroxy derivative: Mycobacterium fortuitum NCAIM 00327, Mycobacterium fortuitum NCAIM 00323 and Nocardia sp. RG 1369.
All of the three strains are able to convert the compound of the formula (II), wherein the meaning of -A-A′- is —CH2—CH2— group, into 9α-hydroxy derivative. However we found, that only Nocardia sp. RG 1369 strain is able to convert the compound of the formula (II), wherein the meaning of -A-A′- is —CH═CH— group, into 9α-hydroxy derivative.
In order to improve this conversion ability we carried out experiments in shaken flasks, using glucose, saccharose or glycerol as carbon source, preferably 5-25 g/dm3 glycerol, more preferably 15 g/dm3 glycerol, as well as using yeast extract, plant peptone or malt extract as nitrogen source, preferably using the yeast extract, the plant peptone and the malt extract together in 1-10 g/dm3 concentration, more preferably in 5-5 g/dm3 concentration, in given case applying ammonium, phosphate, potassium, magnesium and iron in their appropriate compounds. The cultivation temperature was 28-35° C., preferably 32° C. When Nocardia sp. RG 1369 strain was cultured as mentioned above and the canrenone substrate was added in 4 g/dm3 concentration we found that a significant amount of the steroid decomposed in a few hours, although the 9α-hydroxy-canrenone product can still be isolated, but after 24 hours of the addition of the substrate the total degradation of the steroid backbone was observed.
In our further experiments we tried to shift the reaction towards the formation of 9α-hydroxy-canrenone by using a selective inducer. From among the known inducers AD (chemical name: 4-androstene-3,17-dione) and 10,11-dihydroxy-levodione (chemical name: 13-ethyl-10,11α-dihydroxy-4-gonene-3,17-dione) were active. When 10,11-dihydroxy-levodione was used as an inducer the decomposition took place 6-10 hours later, than in the case of AD. The 10,11-dihydroxy-levodione inducer was dissolved in a mixture of methanol-water, preferably in a 3:1 mixture, at elevated temperature, preferably at 50° C. and filtered to obtain a sterile solution. It was added to the culture at the end of the lag period, after 10-24 hours, preferably after 18 hours, in 0.01-0.5 g/dm3 concentration, preferably in 0.05 g/dm3 concentration.
According to our experiments the degradation can be delayed by addition of Δ1-dehydrogenase enzyme inhibitors such as chloramphenicol, oxytetracycline and streptomycin antibiotics as well as quinones, for example hydroquinone, naphthoquinone and ninhydrin. We obtained the best results when we used streptomycin; the decomposition time was 3-7 hours longer. In our experiments the antibiotic was added 2-8 hours after the induction, preferably after 6 hours, in 2-10 mg/dm3, preferably in 6 mg/dm3 final concentration.
After analyzing the results of our experiments we recognized that we have to try to produce a strain starting from Nocardia sp. RG 1369 strain by mutagenic treatment and selection, which can be used in the industrial process for the synthesis of the compound of the formula (I), wherein the meaning of -A-A′- is —CH2—CH2— or —CH═CH— group.
From among the possible mutagenic treatments we choose irradiation with UV light of 254 nm wavelength. During the mutagenic treatment the culture of Nocardia sp. RG 1369, which was suspended in physiological saline and kept under aseptic condition, was treated by known method using Mineralight UVGL-58 type lamp from 15 cm for 23 min—the irradiation time was chosen on the basis of the previously measured lethality curve.
The neat cultures, which were obtained by known methods, were screened and surprisingly it was found, that there was one isolate, which was able to convert the compound of the formula (II), wherein the meaning of -A-A′- is —CH2—CH2— or —CH═CH— group, into the compound of the formula (I), wherein the meaning of -A-A′- is —CH2—CH2— or —CH═CH— group, without considerable degradation. The so obtained Nocardia sp. F1a (RG 4451) bacterium mutant was able to perform higher than 80% conversion. Upon rRNA sequencing the bacterium was identified as Nocardia farcinica NCAIM (P)—B 001342 and deposited on 4 Jun. 2007 for the purposes of patent procedure under the Budapest Treaty at the Hungarian National Collection of Agricultural and Industrial Microorganisms (NCAIM), Budapest, Somloi ut, 14-16, 1118, Budapest, Hungary.
According to the above mentioned facts the invention relates to a process for the selective synthesis of compound of the formula (I),
wherein the meaning of -A-A′- is —CH2-CH2- or —CH═CH— group, by the bioconversion of compound of the formula (II),
wherein the meaning of -A-A′- or —CH═CH— group, comprising using Nocardia farcinica bacterium strain, deposition number of which is NCAIM (P)—B 001342, as hydroxylating microorganism in the bioconversion.
The morphological characteristics of the new mutant Nocardia farcinica NCAIM (P)—B 001342 strain show small dissimilarity to those of the starting Nocardia sp. RG 1369 strain. This difference is most visible on the surface of YTA agar (composition of which is: 10 g/dm3 of tripcasein; 1 g/dm3 of yeast extract; 5 g/dm3 of sodium chloride; 0.25 g/dm3 of magnesium sulfate heptahydrate; 0.07 g/dm3 of calcium chloride dihydrate; 20 g/dm3 of agar-agar): the starting Nocardia sp. RG 1369 strain produces yellow-orange pigment and its surface is plain, shiny, most of the developed culture can be found below the surface of the agar and not above it. In contrast to this the surface of cultures of the new mutant Nocardia farcinica NCAIM (P)—B 001342 strain is not plain, but wrinkled and only small portion of them can be found below the surface of the agar.
The identification of the bacterium strain was done by the partial sequence analysis of 16S rRNA gene.
>RG1(Nocardia sp. F1a (RG 4451) fullseqed2, hereinafter SEQ ID NO: 1:
GTCGAGCGGTAAGGCCCTTCGCGGTACACGAGCGGCGAACGGGTGAGTAA CACGTGGGTGATCTGCCCTGTACTTCGGGATAAGCCTGGGAAACTGGGTC TAATACCGGATATGACCTTACATCGCATGGTGTTTGGTGGAAAGATTTAT CGGTACAGGATGGGCCCGCGGCCTATCAGCTTGTTGGTGGGGTAATGGCC TACCAAGGCGACGACGGGTAGCCGGCCTGAGAGGGCGACCGGCCACACTG GGACTGAGACACGGCCCAGACTCCTACGGGAGGCAGCAGTGGGGAATATT GCACAATGGGCGAAAGCCTGATGCAGCGACGCCGCGTGAGGGATGACGGC CTTCGGGTTGTAAACCTCTTTCGACAGGGACGAAGCGCAAGTGACGGTAC CTGTAGAAGAAGCACCGGCCAACTACGTGCCAGCAGCCGCGGTAATACGT AGGCTGCGAGCGTTGTCCGGAATTACTGGGCGTAAAGAGCTTGTAGGCGG TTTGTCGCGTCGTCCGTCAAAACTTGGGGCTCAACCCCAAGCTTGCGGGC GATACGGGCAGACTTGAGTACTGCAGGGGAGACTGGAATTCCTGGTGTAG CGGTGAAATGCGCAGATATCAGGAGGAACACCGGTGGCGAAGGCGGGTCT CTGGCCAGTAACTGACGCTGAGAAGCGAAAGCGTGGGTAGCGAACAGGAT TAGATACCCTGGTAGTCCACGCCGTAAACGGTGGGCGCTAGGTGTGGGTT TCCTTCCACGGGATCCGTGCCGTAGCTAACGCATTAAGCGCCCCGCCTGG GGAGTACGGCCGCAAGGCTAAAACTCAAAGGAATTGACGGGGGCCCGCAC AAGCGGCGGAGCATGTGGATTAATTCGATGCAACGCGAAGAACCTTACCT GGGTTTGACATACACCGGAAACCTGCAGAGATGTAGGCCCCCTTGTGGTC GGTGTACAGGTGGTGCATGGCTGTCGTCAGCTCGTGTCGTGAGATGTTGC GTTAACTCCCGCAACGAGCGCAACCCTTGTCCTGTGTTGCCAGCGCGTTA TGCCGGGGACTCGCAGGAGACTGCCGGGGTCAACTCGGAGGAAGGTGGGG ACGACGTCAAGTCATCATGCCCCTTATGTCCAGGGCTTCACACATGCTAC AATGGCCGGTACAGAGGGCTGCGATACCGTGAGGTGGAGCGAATCCCTTA AAGCCGGTCTCAGTTCGGATCGGGGTCTGCAACTCGACCCCGTGAAGTTG GAGTCGCTAGTAATCGCAGATCAGCAACGCTGCGGTGAATACGTTCCCGG CCCTTGTACACACCGCCCGTCACGTCATGAAAGTCGGTAACACCCGAAGC CGGTGGCCTAACCCCTTGT
The obtained sequence (1396 bp) covers the 91% of the full gene (1527 bp).
The species identification of the studied strain can be determined on the basis of the NCBI BLAST hits by the applied genotaxonomical method: the correct species designation of RG 4451 strain is Nocardia farcinica. 
Its place in the systematics of the living organisms: Nocardia farcinica Trevisan 1889 [3] Cellular organisms; Bacteria; Actinobacteria; Actinobacteria; Actinobacteridae; Actinomycetales; Corynebacterineae; Nocardiaceae; Nocardia; Nocardia farcinica 
The exact data of the applied NCBI BLAST [2] identification:
Accessibility: (ncbi.nlm.nih.gov/blast/,) Version: BLASTN 2.2.16 (Mar-25-2007)
Database: All GenBank+EMBL+DDBJ+PDB sequences but no EST, STS, GSS, environmental samples or phase 0, 1 or 2 HTGS sequences); 5,284,371 sequences;
20,692,750,832 total letters, Algorithm: megablast
The 16S rRNS gene sequence of strains belonging to Nocardia farcinica species is identical or very similar to each other. The similarity is also considerable in the subgenus, but the families of Corynebacterineae genus (see Nocardiaceae, Mycobacteriaceae) are very different.
An important and clearly observable difference between the two strains is the conversion ability in synthesizing the compound of the formula (I), wherein the meaning of -A-A′- is —CH2—CH2— or —CH═CH— group, that is the new mutant Nocardia farcinica NCAIM (P)—B 001342 strain retained the 9α-hydroxylation ability, but the degradation of the steroid backbone is suppressed. Therefore—under the previously defined experimental conditions—due to the suppressed degradation the amount of the 9α-hydroxy product is higher, it can be isolated: it can be used on industrial scale synthesis.
The invention is illustrated by the following not limiting examples.